1. Field of the Invention
The present invention relates to a technology for eliminating noise appearing on a background of an image when processing the image.
2. Description of the Related Art
Typically, when outputting image data having a specific color space and a specific data format from a first device to a second device, the color space and the data format of the image data are converted to those that are acceptable to the second device. By using this technique, image data stored in a laser printer (LP) or a multifunction product (MFP) on a network can be transmitted to some other printer on the same network and get printed there. Moreover, image data stored on a certain device can be distributed to other devices, or can be processed by using a personal computer (PC). Therefore, possibilities have increased that a color space for monitor display (hereinafter, “monitor color space”) is widely employed.
Sometimes data for monitor display (hereinafter, “monitor data”) is required to be converted into data for printing (hereinafter, “print data”). When doing so, the data format and the color space are converted into compatible format and space. To adjust the color space between the monitor data and the print data, some color conversion methods are prepared and an appropriate one of the color conversion methods is employed for each object, or an appropriate processing is performed on each of the objects to obtain both better tone and sharpness of the image. Because attribute information is necessary for selecting an appropriate color conversion method, images are stored in combination with corresponding attribute information.
For example, Japanese Patent Application No. 2000-148973 discloses an image processing system including a printer, in which image data is generated in combination with attribute information for each pixel of the image, and the image data is processed according to the contents of the attribute information before the image is printed out. Specifically, the printer switches processing methods for converting image data to output a high-quality image.
Because both of the image data and the attribute information need to be stored for each pixel, it is desirable that a size of the data should be as small as possible. The typical monitor color space corresponds to RGB (red, greed, and blue) 8-bit data. For example, when the image data is stored by using the color space and data format corresponding to the monitor color space, the data size becomes three times an image size. If the printing resolution is, for example, 600 dots per inch (dpi) in the A-3 size, a large memory capacity of about 198 mega bytes (Mbytes) is required for the image. Therefore, an image data compression is essential for storing the image data, and a lossy compression method is more suitable than a lossless compression method to such large data.
While some data loss is caused with the lossy compression method, the lossy compression method greatly increases the data compression ratio. For this reason, the lossy compression method is widely used to compress a still image, a video, and audio data, which is not greatly influenced by data loss or a change in the data not as in the case of computer programs or character data, realizing a highly-efficient data compression.
In an image compressed by using the lossy compression method, however, noise sometimes appears on the image in an area, which corresponds to a white area (or the single-color background) of the original image, even with an adjustment of a compression parameter. A color sometimes appearing on the white area (or the single-color background) of the original image, for example, around a character, is visually noticeable. Although the lossless compression method that causes no data loss can be used instead, high compression efficiency cannot be obtained compared to the lossy compression method.